A run on a bank occurs when depositors withdraw deposits quickly. In the days before deposit insurance, bank runs from retail depositors were common as their money was genuinely at risk from bank failure. These runs were damaging as banks which were solvent but not sufficiently liquid (due to demand deposits funding term loans) could fail due to a run. These days, bank runs occur in the wholesale funding market: hence Gorton and Metrick’s influential paper about the run on repo in the 2008 crisis. These repo runs occur if wholesale funders refuse to carry on providing liquidity.
Hence a run on a CCP can happen if the CCP’s counterparties withdraw liquidity quickly, and the CCP cannot liquidate assets fast enough to keep up with the liquidity demands on it (or can only do so at great cost).
What’s different about CCPs is that this can happen in two ways.
- If clearing members neutralise their risk quickly, CCPs have to repay initial margin. That initial margin might have been provided as cash but invested by the CCP in securities. Therefore this kind of run – a run on IM – occurs when the CCP cannot liquidate investment assets fast enough to repay IM. CCP investment mandates are designed to reduce the risk of this happening. Of course, one way that this can happen is if clearing members move their risk to another CCP. If clearing member house portfolios are regularly compressed, then moving the house account may not be that difficult a matter. Hence the risk of a large clearing member moving its risk from one CCP to another clearing the same products is non-zero.
(One could imagine this happening for instance because confidence was waning the CCP, or because the CCP’s ‘qualifying’ status was in doubt, leading to the possibility of higher capital requirements for its users.)
- A less obvious but equally risky run mechanism (a route for a run?) is collateral substitution. As yield curves normalise over the coming years, the incentive to hold margin in the form of securities vs. cash will change. Term securities may look more attractive, resulting in clearing members substituting securities for cash in margin accounts. This can create liquidity pressure too as the CCP has to liquidate what it has invested the cash in to refund it. There is a case in this context for limiting the speed at which clearing members can substitute non-cash for cash margin.
Time will tell if these risk vectors do indeed threaten CCPs. But until we know more, alertness to potential CCP liquidity risks is vital.